PROJECT SUMMARY Protein-ligand binding events underlay all life processes. Protein design tests and extends our knowledge of protein folding and function through the creation of proteins from scratch. This proposal aims to develop a computational method for the design of proteins that bind to any small molecule with high affinity and selectivity. The state-of-the-art in ligand-binding protein design critically relies on random experimental optimization and screening. If we truly understand how proteins bind small molecules, we should be able to go directly from computer models to tight binders. The hypothesis that drives this proposal is that proteins use a vast but now enumerable number of molecular interaction motifs combinatorially throughout evolution to create the binding sites of modern-day proteins. Computational methods will be employed to uncover this set of interactions in the large database of protein structures available in the protein databank (PDB). Binding sites will be designed by sampling motifs for all functional groups of a ligand onto a protein backbone. We call this design method Convergent Motifs for Binding Sites (COMBS). COMBS was used to design ABLER, the first ligand-binding protein designed from scratch to bind its target ligand?the antithrombotic drug apixaban?with an unprecedentedly high affinity, without experimental optimization of sequence. ABLER has potential clinical relevance as an anti-clotting antidote, although that is outside the scope of the proposal. High-resolution crystal structures of ABLER agree with the design model, both in overall topology and the intended molecular interactions with the ligand. Aim 1 of this proposal focuses on designing variants of ABLER to increase affinity and probe the molecular bases for the observed drug-protein interactions. Aim 2 focuses on the role of water in ligand-binding protein design, motivated by the water-mediated protein-ligand interaction found in the crystal structure. In this Aim, I will curate a database of water-protein interactions from the PDB and use these to sample water-mediated protein-ligand interactions during design. I will also learn to use explicit-water molecular dynamics simulations to critically assess the roles of water in binding. Aim 3 uses COMBS to redesign the binding site of pyrrolysine tRNA synthetase for incorporation of charged unnatural amino acids (such as sulfotyrosine) into mammalian cells, since laboratory evolution and library screens for this goal have so far been unsuccessful. These aims will augment my training in molecular biology, computational protein design, and protein structural characterization (X-ray crystallography and NMR). The K99 portion of this work in the DeGrado lab will expose me to all aspects of the scientific process, from inception to publication. Bill is a world expert in protein design, and his insight is critical to the success of the project. At UCSF, I will gain much through my regularly scheduled meetings with Ethan Weiss, who brings the perspective of a physician scientist with a long history of antithrombotic research and clinical applications. My collaboration with UCSF professor Lei Wang will expose me to the field of unnatural amino acid incorporation and will be critical for applying COMBS to the most impactful targets for mimics of post-translational modifications. The research and training proposed herein will greatly complement my current skillset and background, which will be essential to my research program as I transition into an independent principal investigator.